Congress’ investigative arm is spotlighting the challenges states face with their testing systems, from recession-driven budget pressures to ensuring the validity and reliability of statewide student assessments.
A new, wide-ranging report by the Government Accountability Office on state testing under the No Child Left Behind Act finds that although most states today spend far more on assessments than in 2002, when the federal law was enacted, 19 reported recent cuts in their testing budgets because of fiscal constraints. Ten others expected future reductions in such spending.
The report also finds that states face a variety of hurdles in ensuring the validity and reliability of those tests, such as staff capacity, assessment security, and developing alternate assessments for students with disabilities.
In addition, the GAO says, some states are using more multiple-choice items than at the time of the NCLB law’s enactment because they can be scored inexpensively within the tight reporting time frames of the federal law.
“Cost and time pressures have influenced state decisions about assessment type … and content,” says the GAO report, issued last month.
John R. Tanner, the director of the Center for Innovative Measures at the Council of Chief State School Officers, said he sees real risks to an over-reliance on multiple-choice items.
“One of the consequences of test-based accountability is that, particularly in low-performing schools, the test becomes the curriculum, because passing that test becomes so all-important,” said Mr. Tanner. “If you don’t have items that represent the full richness of the [academic] standards, … then you put the curriculum at risk of doing the same thing.”
Meanwhile, nearly eight years after the law’s passage, a dozen states as of July still had not received final approval for their reading/language arts and mathematics assessment systems under the NCLB act by the U.S. Department of Education, according to the GAO. One key difficulty in most of those states, the report says, was ensuring the validity and reliability of their alternate assessments.
To win federal approval, states must show evidence of sufficient technical quality in their assessments and alignment with their academic standards.
The GAO report also raises concerns about the review process the Education Department has used to inform its determinations.
The No Child Left Behind Act, a reauthorization of the Elementary and Secondary Education Act, ramped up the extent of state tesing. It requires all states to test students in reading/language arts and math each year from grades 3-8, and once in high school.
Data from the assessments are used to make accountability decisions mandated under the law, including corrective actions such as replacing some school staff. The law also requires science testing once in elementary school, middle school, and high school. And, under the NCLB act, states must develop alternate assessments for students with significant cognitive disabilities.
The federal government has provided $400 million each year to help states pay for NCLB-mandated testing. But the GAO report suggests this often fails to cover the full costs.
More than half the 44 states that responded to a GAO survey question reported that the majority of their funding for tests administered to meet NCLB requirements came from state coffers.
As of January, 19 states told the GAO that they had reduced their testing budgets to comply with NCLB mandates as a result of statewide budget cuts. Another 14 states said their budgets for NCLB testing had not been reduced, but 10 of those expected future cuts.
States are taking a range of steps to trim testing budgets, both for NCLB-required tests and beyond.
“They’re cutting some of the technical staff” at state education agencies, said Scott Marion, the associate director of the Center for Assessment, a nonprofit consulting firm based in Dover, N.H. “Certainly, there is more multiple choice [material]. They’re cutting out tests that they might have included before, like social studies.”
Ohio, for one, recently suspended its social studies tests for 5th and 8th graders, and its writing tests for 4th and 7th graders. Those tests were selected because they are not required under the NCLB law, said Scott Blake, a spokesman for the Ohio education department. The test suspensions will save the state $8.2 million over two years, he said.
States also have cut costs by increasing their use of multiple-choice questions on their tests since the federal law was enacted, the GAO found. Of the 47 states that responded to a GAO question on the matter, 10 reported increasing the use of multiple-choice items on reading/language arts tests, and 11 reported an increase in math. Marland, for example, previously administered an assessment that was almost totally comprised of “open/constructed response questions,” but has since moved to an assessment that is primarily multiple choice.
Even so, the report states, “Many state officials and experts we spoke with told us that multiple-choice items limit states from assessing highly cognitively complex content. For example, Texas assessment officials told us that some aspects of state standards, such as a student’s ability to conduct scientific research, cannot be assessed using multiple choice.”
At the same time, the GAO found that some states have developed and administered assessments in collaboration with one another, allowing them to pool resources and use a greater diversity of item types. It cites as an example the New England Common Assessments Program, which involves Maine, New Hampshire, Rhode Island, and Vermont.
‘Gaps’ in Security
Meanwhile, the report finds that the capacity of state education agencies to develop and oversee testing systems varies widely.
“Greater state capacity allows states to be more thoughtful in developing their state assessment systems, and provide greater oversight of their assessment vendors, according to state officials,” the GAO says.
The report identifies several gaps in security policies for assessments. For one, experts told the GAO that many states do not conduct any statistical analyses of test results to detect indications of cheating.
The report also finds that states have faced difficulties in developing valid and reliable alternate assessments for students with disabilities. One problem, the GAO says, is that this student population is extremely diverse, with a wide range of cognitive disabilities.
Finally, the report raises concerns about the quality of communication from the federal Department of Education and the independent peer-review panels it established to scrutinize state assessment systems, suggesting it’s been decidedly mixed.
Gregory J. Cizek, a professor of educational measurement and evaluation at the University of North Carolina at Chapel Hill, said poor communication from the department and peer-review panels has been an “epidemic” problem for states.
“States were kind of groping to figure out what would be acceptable,” he said. “That has caused a lot of variability, ... where two states [may be] doing the same thing, and one gets approved and another doesn’t.”
In comments submitted to the GAO, the Education Department said it has taken recent steps to improve communication with states during the peer-review process.
A version of this article appeared in the October 14, 2009 edition of Education Week as States Face Budget Hurdles on Testing, GAO Says